Geochemistry, Geophysics, Geosystems

Magnetic fabric of stalagmites and its formation mechanism

Authors


Abstract

[1] Speleothems have been regarded as ideal archives for recording variations of the Earth's magnetic field because they crystallize rapidly and are seldom modified by post-depositional processes, and because they can be precisely dated. Magnetic fabric research on speleothems has potential to reveal details of their interior structure and the distribution of ferrimagnetic minerals, which would benefit investigations of geomagnetic field behavior. Two stalagmites (HS4 and WD1) from caves in the middle reaches of the Yangtze River were collected for magnetic fabric study. The anisotropy of magnetic susceptibility (AMS) of the stalagmites is mainly controlled by calcite crystals and the directions of the minimum axes of AMS ellipsoids indicate that their crystallographic orientations are roughly perpendicular to the growth laminae. Anisotropy of the isothermal remanent magnetization (AIRM) indicates that the distributions of ferrimagnetic minerals in the stalagmites are not correlated with stalagmite growth laminae. Mean directions of the maximum principal axes of AIRM (R1 axes) for WD1 are close to that of the natural remanent magnetization, which suggests that the orientation of ferrimagnetic minerals in this stalagmite are likely to have been controlled by the geomagnetic field.

1. Introduction

[2] The anisotropy of magnetic susceptibility (AMS) and anisotropy of isothermal remanent magnetization (AIRM) are physical properties of rocks that can be used for petrofabric and structural studies [Tarling and Hrouda, 1993; Borradaile and Henry, 1997; Herrero-Bervera et al., 2001; Lagroix and Banerjee, 2002, 2004; Wing and Ferry, 2007]. Generally, AMS is controlled by the ferrimagnetic, paramagnetic and diamagnetic minerals, while AIRM is controlled by ferrimagnetic minerals [Jackson, 1991].

[3] Speleothems are terrestrially deposited carbonate rocks. They crystallize rapidly, are seldom modified by post-depositional process, and they can be dated precisely using U-Th dating and layer counting, therefore, they have potential for paleomagnetic studies of geomagnetic field behavior [Latham et al., 1979, 1989; Morinaga et al., 1992; Perkins and Maher, 1993; Lean et al., 1995; Openshaw et al., 1997]. Lascu and Feinberg [2011] recently reviewed paleomagnetic and rock magnetic work on speleothems and concluded that speleothems are ideal archives for investigating Earth's magnetic field. The magnetic fabric of speleothems can reveal details of their interior structure and the distributions of magnetic minerals, which will benefit paleomagnetic investigations of speleothems. However, to the best of our knowledge, no such work has been reported.

[4] In this study, two stalagmites were collected from caves in the middle reaches of the Yangtze River, China, to determine the AMS and AIRM characteristics of stalagmites and to better understand the mechanisms by which magnetic fabrics form these rocks.

2. Samples and Location

[5] The HS4 stalagmite is from Heshang cave and WD1 is from Wudi cave in the middle reaches of the Yangtze River. HS4 is 2.5 m long, with a diameter of 28–38 cm. It was actively growing when recovered in 2001. U-Th dating indicated that its basal age is 9.52 ka B.P. [Hu et al., 2008]. HS4 is gray and relatively pure calcite with clear annual bandings (Figure 1). WD1 is conically shaped and is 47 cm in height and 8–27 cm in diameter. It was actively growing when collected in 2003. This stalagmite has not been dated. WD1 is pure and dense with a gray to slightly pink color. Clear growth rings can be observed from cross-sections through the stalagmite (Figure 1). The growth rings in WD1 and annual bandings in HS4 appear as alternate bright and dark laminae with different widths.

Figure 1.

Schematic sections and sampling positions for stalagmites HS4 and WD1. Samples (2 × 2 cm cubes) from HS4 were cut from the central part of HS4 (in the rectangular frames) where the bandings (solid lines) are horizontally distributed. Samples (cylinders with 2.5 cm diameter and 2.2 cm height) from WD1 were drilled from three circular cross-sections through the stalagmite. Sections from left to right are from the top, middle, and bottom of WD1, respectively. Dotted circles are drilling positions, and the number in each circle denotes the sample number. The artificial north direction (N) was labeled to provide uniform orientation for all samples. Directions of sample x axes are parallel to the “N” direction. Directions of sample z axes in HS4 are indicated by an arrow, and those in WD1 are vertical downward with respect to the cross-sections.

[6] The stalagmites were not oriented when collected from the caves, but each stalagmite was given an artificial direction (labeled as “N”) to make sure that each sub-sample from the same stalagmite has the same direction. 188 oriented samples (99 from HS4 and 89 from WD1) were obtained and labeled with z and x axes. The z axes are vertically downward (pointing from the stalagmite's top to its root, see detail inFigure 1), and the x axes are parallel to the artificial north direction (labeled as “N”). For HS4, samples (2 × 2 cm cubes) were cut from the central part of the stalagmite where the bandings are horizontally distributed (Figure 1), so they all have horizontal growth laminae. For WD1, samples (cylinders with 2.5 cm diameter and 2.2 cm height) were drilled from three circular cross-sections through the stalagmite. Their positions are shown inFigure 1. These samples have curved growth laminae with orientation that can be measured as follows: (1) with the samples placed horizontally and their artificial north direction parallel to geographic north, the declination of the average growth laminae were measured with a protractor, and (2) the angle of inclination was measured with a protractor as the angle of the laminae with respect to horizontal. Some unoriented sub-samples were prepared for chemical and magnetic mineral analysis. All samples were cleaned ultrasonically 4 times (15 min each time) with Milli-Q water to remove all adhering particles.

3. Analytical Techniques

3.1. Chemical Analysis

[7] Previous research has demonstrated that impurities (cations other than calcium) can affect the AMS of carbonate rocks, including substitution of Fe2+ and Mn2+ for Ca2+, which makes normally diamagnetic carbonate minerals more paramagnetic [Schmidt et al., 2006; Almqvist et al., 2010; Borradaile and Jackson, 2010]. In order to provide information on the purity of the studied stalagmites, elemental compositions of the stalagmites were determined by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). A 193 nm Ar-F (GeoLas 205M, Microlas) excimer laser system was used for the ablation, in combination with a 7500a ICP-MS (Agilent Technologies) instrument, which was run under the same conditions as those reported byGuo et al. [2008]. Six samples from HS4 and two samples from WD1 were prepared as thin sections with thicknesses of about 2 mm. Each sample was ablated at 5–6 points, of which 2–3 points were distributed in the bright laminae with the others in the dark laminae. The diameter of the laser beam was 80 nm and every point was ablated 400 times with a 100 nm depth, and then average concentrations were calculated. Synthetic CaCO3 standards were doped with trace elements of interest using a coprecipitation method. These samples were used as the matrix matching standard [Guo et al., 2008].

3.2. AMS Measurements

[8] Magnetic anisotropy was measured at room temperature with a KLY-4S magnetic susceptibility meter (AGICO, Brno), using an alternating field of 300 A/m and a 875 Hz measurement frequency. AMS is defined as a second rank tensor that is usually visualized as a susceptibility ellipsoid. The shape of the susceptibility ellipsoid is defined by three perpendicular principal axes,κ1, κ2 and κ3 (the maximum, intermediate and minimum axes of the magnetic susceptibility tensor, respectively).

3.3. Magnetic Mineral Analysis

[9] The natural remanent magnetization (NRM) and a saturation isothermal remanent magnetization (SIRM, imparted in a field of 1000 mT) of the oriented samples were measured using a JR-6A spinner magnetometer to investigate the characteristics of ferrimagnetic minerals in the stalagmites. Six representative unoriented samples were magnetized in stepwise increasing DC fields to obtain IRM acquisition curves, using 19 steps from 10 to 1200 mT. After imparting the IRM at 1200 mT, back-field demagnetization was carried out to determine the coercivity of remanence (Bcr) for each sample. Ten samples (5 from HS4 and 5 from WD1) were subjected to thermal demagnetizaton analysis of a three-component IRM [Lowrie, 1990]. The samples were first magnetized in a field of 1000 mT along their z-axes, which magnetizes most of the ferrimagnetic minerals in the direction of the maximum field. Next, a field of 300 mT was applied along the y-axes, thus remagnetizing the coercivity fraction softer than 300 mT along the y-axes and leaving the higher coercivity minerals magnetized along the sample z-axes. Finally, a field of 100 mT was applied along the sample x-axes. The samples were then demagnetized thermally in a TD-48 thermal demagnetizer.

3.4. AIRM Measurement

[10] In order to reveal the distribution of ferrimagnetic minerals in samples [Jackson, 1991], 81 samples, including 40 from HS4 and 41 from WD1, were selected randomly for AIRM determination. The AIRM of the chosen samples was determined after measurement of the NRM but before the SIRM measurement mentioned in section 3.3. The AIRM determination included three steps: (1) the sample was demagnetized in an alternating field (AF) with 100 mT peak field and 0.0075 mT/half-cycle decay rate (D-2000 AF Demagnetizer); (2) the sample was magnetized along a designed direction with a direct magnetic field of 30 mT for stalagmite HS4 and 10 mT for WD1 (IM-10-30 Impulse Magnetizers); and (3) the IRM was measured with a JR-6A spinner magnetometer. Applied direct fields of 30 mT and 10 mT were the weakest magnetic fields that could be applied to create a detectable IRM for HS4 and WD1, respectively. Each of the above steps was repeated 6 times, and each time the samples were magnetized along different directions. According to the model B recommended byJelínek [1993], the six magnetizing directions are: declination (d) = 45° and inclination (i) = 0°, d = 315° and i = 0°, d = 90° and i = 45°, d = 90° and i = −45°, d = 360° and i = 45°, and d = 180° and i = 45°. The application of putting 6 magnetizing positions of a same sample in a fixed position of pulse magnetizer can reduce the influence of possible inconsistence of the imparting magnetizing field. Finally, the AIRM tensor values were calculated based on the 6 groups of IRM data using the AREF program (AGICO).

3.5. Crystal Texture Analysis

[11] The preferred crystallographic orientation of calcite in the stalagmites was measured using the electron backscatter diffraction (EBSD) technique with a Quanta 2000 scanning electron microscope. Two thin sections were doubly polished and oriented, with the sample xaxis perpendicular to the growth laminae. The measurement was carried out with an accelerating voltage of 20 kV, a spot size of 6.5 and a working distance of 25 mm. Diffraction patterns were manually (or automatically for mapping mode) collected and indexed using the HKL Channel 5+ software. To assure data quality, only those measurements with mean angular deviation values below 1° were used further. The orientation of the pole of the [001] plane (i.e., c-axis) of the calcite crystal was grouped for each sample.

4. Results

4.1. Chemical Composition

[12] The results of chemical analyses are summarized in Table 1. The stalagmites have Ca concentrations of 386700 ppm (96.7% of CaCO3) in HS4 and 398000 ppm in (99.5% of CaCO3) WD1. The concentrations of impurity elements (other than Ca) ranged from 6942 to 12880 ppm in HS4, and are much lower in WD1, from 1178 to 1352 ppm. Mg is the major impurity in HS4 with concentrations >6000 ppm. Na, Si, Ba, and Sr are secondary impurities with concentrations <500 ppm. Mg, Al, Si and Sr are important impurities in WD1 with concentrations ranging from 40 to 580 ppm. The concentrations of other impurity elements, such as K, Ti, V, Cr, Ni, Cu, Zn, Sn, and Pb, are generally <10 ppm. The total concentrations of Fe (<60 ppm) and Mn (<1 ppm) in both of the stalagmites are extremely low, and they may not be derived solely from the substitution of Fe2+ and Mn2+ for Ca2+.

Table 1. Chemical Composition of Stalagmites From LA-ICP-MS Analysesa
Sample CaMgNaAlSiKTiVCrMnFeNiCuZnSrSnBaPb
  • a

    Units are ppm. Samples HS4–1 to HS4–6 are located 13, 84, 132, 148, 195 and 236 cm from the top of HS4, respectively. WD1–1 and WD1–2 come from the middle cross-section of WD1.

HS4–1Average3904986114340.72823.40.20.50.60.7240.90.54.81090.33660.1
SD21791452160.71605.30.20.40.70.960.20.12.0170.1700.1
HS4–2Average3868928412341.14323.20.20.10.50.1240.70.35.71140.44870.1
SD1762117981.2411.90.2<0.10.60.240.20.25.4140.2690.1
HS4–3Average3875917834340.74925.6<0.10.10.30.4201.40.43.91000.24800.1
SD16991120170.9774.7<0.10.20.50.460.90.22.4120.161<0.1
HS4–4Average3873248208260.73662.90.20.20.60.9221.10.54.61070.3497<0.1
SD3678264860.61461.30.20.20.20.620.40.31.3300.1163<0.1
HS4–5Average3857369222390.74335.90.40.10.40.8241.40.34.81285.64860.3
SD42403002130.5722.90.40.10.40.630.60.13.52713.31070.3
HS4–6Average38227811839431.73294.50.1<0.10.31.2371.40.54.51300.34880.1
SD3911257481.41152.30.2<0.10.41.6110.70.31.5240.1105<0.1
WD1–1Average39789437058257511.21.80.20.31.2591.40.10.32360.280.1
SD7091264702229.61.60.10.11.6410.30.20.3190.03<0.1
WD1–2Average3981363903485055.01.60.20.60.7460.50.20.11710.25<0.1
SD384871222271.91.00.10.50.7120.80.10.160.11<0.1

4.2. AMS Characteristics

[13] The studied stalagmites are diamagnetic with negative values for the three principal susceptibilities (κ3 ≤ κ2 ≤ κ1 < 0). The mean susceptibility is the arithmetic mean of the eigenvalues (Кm = (κ1 + κ2 + κ3)/3). The susceptibility difference is defined as ΔК = κ1κ3. The degree of anisotropy (Pa), the magnetic lineation (La), and the magnetic foliation (Fa) were calculated using the absolute values of the principal susceptibilities (Кa1 = |κ3| ≥ Кa2 = |κ2| ≥ Кa3 = |κ1|) [Hrouda, 2004] to allow these values to exceed 1. The parameters are then: Pa = Кa1a3 = |κ3|/|κ1|, La = Кa1a2 = |κ3|/|κ2|, and Fa = Кa2a3 = |κ2|/|κ1|. The average values of these AMS parameters are listed in Table 2.

Table 2. Average Values of AMS Parameters for Stalagmites HS4 and WD1a
StalagmiteКm (SI)χ (m3/kg)ΔК (m3/kg)LaFaPa
  • a

    The average density is 2.03 × 103 kg/m3 for HS4 and 2.59 × 103 kg/m3 for WD1.

HS4−8.8E-6−4.2E-93.4E-101.071.011.08
WD1−11.2E-6−4.3E-93.8E-101.081.011.09

[14] The average mass susceptibility (χ) values are similar for HS4 and WD1, but the distribution of values for WD1 is more negative and narrower than HS4 (Figures 2b and 2c). The Pa and Δκ values indicate that all measured samples have a strong AMS (Table 2). In addition, Δκ and Pa do not correlate with the mass susceptibility (Figures 2b and 2c). The average La value is much higher than Fa (Table 2). In Flinn-type diagrams (Figure 2a), the data points are distributed close to the La axis, which suggests that the susceptibility ellipsoids are prolate.

Figure 2.

Distributions and correlations of AMS parameters for the studied stalagmite. (a) Flinn-type diagram, in which all of the data points lie close to the La axis, which indicates prolate susceptibility ellipsoids. (b) χ-ΔК and (c)χ-Pa diagrams in which χ values for HS4 are more scattered than those for WD1, and the values of ΔК and Pa are not significantly correlated with the mass susceptibility.

[15] An equal-area stereographic projection of the principal susceptibility axes for HS4 indicates that theκ1 and κ2 axes are distributed within the plane of the stalagmite laminae, while κ3 axes are distributed perpendicular to the laminae (Figure 3). For WD1, there is no distinct distribution pattern of the κ1 and κ2 axes (Figures 4b, 4e, and 4h). However, the distributions of the κ3 axes match well with the sampling locations (Figure 1); that is, every κ3 data point was located perpendicular to the corresponding laminae inclination. After 100% untilting, according to the average orientation of the growth laminae, the principal susceptibilities for WD1 have a similar distribution pattern as the HS4 samples (Figure 3), i.e., κ1 and κ2 are distributed in the plane of the laminae and κ3 is distributed perpendicular to the laminae (Figures 4c, 4f, and 4i). This again indicates that the κ3 axes are perpendicular to the growth laminae.

Figure 3.

Equal-area stereographic projections (lower hemisphere) of the principal susceptibility axes for stalagmite HS4. The projection plane is parallel to the growth laminae. Theκ3 axes are perpendicular to the growth laminae, while κ1 and κ2 are parallel to the growth laminae.

Figure 4.

Equal-area stereographic projections (lower hemisphere) of principal susceptibility axes for the three cross-sections through stalagmite WD1 (seeFigure 1). (a, d, g) Projections of the κ3 axes before untilting of the laminae; (b, e, h) projections of the κ1 and κ2 axes before untilting of the laminae; (c, f, i) projections of the κ1, κ2 and κ3 axes after untilting of the laminae.

4.3. Ferrimagnetic Minerals

[16] The average NRM intensity is 7.10E-05 A/m for HS4 and 1.76E-04 A/m for WD1, and the average SIRM intensity is 8.26E-03 A/m for HS4 and 5.13E-03 A/m for WD1, which indicates the presence of ferrimagnetic minerals in both stalagmites. IRM acquisition curves for HS4 rise steeply at the beginning and 95% of the saturation IRM is reached at 300 mT. Back-field demagnetization curves indicate that Bcr values are lower than 35 mT. Each component of the composite IRM undergoes major reduction at 350°C (Figure 5), which indicates that the soft, medium and hard magnetic fractions in HS4 are caused by the same ferrimagnetic minerals. The present evidence is insufficient to identify these magnetic minerals. For example, Roberts and Pillans [1993]demonstrated that unblocking of three-axis thermal demagnetization curves in the 300–400°C range can result from the presence of a number of minerals, including maghemite, titanomagnetite and magnetic iron sulfides. So such data can be highly ambiguous. Magnetic iron sulfides are unlikely in such environments, however, titanomagnetite [Latham et al., 1989] and maghemite [Herries and Shaw, 2011] have been identified in the detrital fraction of some speleothems.

Figure 5.

Isothermal remanent magnetization (IRM) acquisition and back-field demagnetization curves for selected stalagmite samples. The IRM is normalized by the IRM imparted at 1200 mT.

[17] IRM acquisition curves for WD1 climb less steeply than for HS4 and about 90% of the normalized IRM is achieved at 300 mT, with Bcr values of 35–45 mT. In Figure 6, the hard, medium and soft components of WD1 all unblock at 650°C, which indicates the presence of hematite. There is also a reduction at 585°C of the soft fraction, which indicates the presence of magnetite in WD1.

Figure 6.

Thermal demagnetization of a three-component IRM produced by magnetizing the sample in a 1000 mT field along its z axis, followed by a 300 mT field along they axis, and finally by a 100 mT field along the x axis. See text for discussion.

4.4. AIRM Characteristics

[18] The distribution of ferrimagnetic minerals was detected using AIRM results. The three orthogonal principal remanent magnetization axes are marked as the maximum (R1), intermediate (R2) and minimum (R3) axes, respectively. The degree of anisotropy of IRM, PIRM = R1/R3, the magnetic lineation, LIRM = R1/R2 and the magnetic foliation FIRM = R2/R3, were calculated according to the formulae for AMS parameters and are listed in Table 3. Both HS4 and WD1 have strong anisotropy of IRM with average PIRM values of 1.18 and 1.43, respectively. However, the AIRM ellipsoids for HS4 are not perfect prolate ellipsoids because of their similar LIRM and FIRM values, while LIRM values for WD1 are notably higher than FIRM values, which indicates a perfect prolate AIRM ellipsoid (Table 3).

Table 3. Average Values of AIRM Parameters for HS4 and WD1
StalagmiteLIRMFIRMPIRM
HS41.091.081.18
WD11.351.061.43

[19] Most R1 axes from HS4 samples cluster in the northwest quadrant of an equal area stereographic projection with R2 and R3 axes roughly distributed in a broad great circle (Figure 7a). R1 axes from WD1 are concentrated in the northeast quadrant (Figure 7b) in geographic coordinates, and the R2 and R3 axes are distributed in a narrow great circle girdle (Figure 7b). After 100% untilting of the laminae, all directions of the principal AIRM axes for WD1 are much more dispersed (Figure 7c). These results suggest that the distribution of R1 axes has no direct relationship with the growth laminae. The AIRM has a distinctly different distribution from the AMS (Figures 3 and 4), which indicates that the distribution of ferrimagnetic minerals does not dominate the AMS of the studied stalagmites.

Figure 7.

Equal-area stereographic projections (lower hemisphere) of AIRM principal axes for HS4 and WD1. (a) An equal-area projection of the AIRM principal axes in HS4; (b and c) equal-area projections of AIRM principal axes in WD1 before and after untilting of laminae, respectively; (d and e) mean directions of the R1 axes and NRM in geographic coordinates for HS4 and WD1, respectively. The “N” direction labeled for each stereographic projection is an artificial direction and is not the real geographic north, as explained insection 2.

[20] The mean direction of the R1 axes from WD1 is close to that of the NRM (Figure 7e), which indicates that the distribution of ferrimagnetic minerals in WD1 might be controlled by the paleomagnetic field. However, this is not the case for HS4 whose mean R1 direction lies far from the mean NRM direction (Figure 7d). The cause of this difference is unclear, but the NRM directions for HS4 are not well clustered (k = 2.7 and α95 = 22.3) (Figure 7d), so the reliability of the NRM signal in this stalagmite is doubtful.

4.5. Crystal Orientation

[21] The carbonate crystals that dominate the stalagmites can be readily recognized by visual inspection. Under the microscope, the long axes of the calcite crystals are almost perpendicular to the growth rings (Figure 8). This finding is similar to previous descriptions [Kendall and Broughton, 1978; Kendall, 1993]. EBSD measurement can reveal crystallographic preferred orientations of calcite crystals. Measurements of the pole to the [001] plane of calcite crystal from EBSD are shown in Figure 9. These results reveal that the poles are mainly distributed along the xaxis, i.e., perpendicular to the growth rings of the stalagmites. Considering that the direction of the pole to the [001] plane is the c-axis of the calcite crystal, the EBSD results indicate that the calcite crystals grew perpendicular to the growth laminae.

Figure 8.

Optical micrographs of thin-sections of stalagmites HS4 and WD1. The sections were of roughly perpendicular to the growth laminae. Micrographs were taken with a Carl Zeiss Axio imager A1m cross Nicols. HS4 has smaller calcite crystals than WD1. The long axes of most crystals lie perpendicular to the growth rings (dashed lines).

Figure 9.

Pole density figures of the [001] plane for the calcite crystals in (a) HS4 and (b) WD1. The pole of the [001] plane for calcite crystals is parallel to the c-axis of the calcite crystal. Thexaxis in the figure lies perpendicular to the growth laminae. This figure illustrates that the poles of the [001] plane of calcite crystals in HS4 and WD1 lie roughly parallel to the x axes, and that mean c-axes are perpendicular to the growth laminae of the stalagmites.

5. Discussion

[22] Stalagmites HS4 and WD1 have weak diamagnetic susceptibility. Small quantities of paramagnetic and ferrimagnetic inclusions can therefore dramatically change the AMS properties of stalagmites. Fe and Mn ions are the major sources of paramagnetism in diamagnetic minerals. Schmidt et al. [2006] demonstrated that substitution of Fe2+ for Ca2+ by more than 500 ppm would produce an inverse magnetic fabric. Borradaile and Jackson [2010] suggested that when ferrimagnetic minerals are present in quantities of ≥0.5% by mass their positive susceptibilities would outweigh the negative diamagnetic susceptibility. Levi and Weinberger [2011] reported that the iron content of the Bar Kokhba limestone was less than 130 ppm, and that there was no dependency between ΔК or P′ and the Fe and Mn contents. In the present study, although the detected Fe and Mn ions result not only from substitution of Fe2+ and Mn2+ for Ca2+, which makes calcite crystals more paramagnetic, they also result from the presence of ferrimagnetic minerals. The total mass percentages of Fe and Mn are far less than 0.01% (Table 1). Therefore, the paramagnetism of Fe and Mn ions in the stalagmites can be neglected. Ferrimagnetic minerals may also dominate the susceptibility [Hrouda et al., 2000]. However, the distributions of ferrimagnetic minerals detected by AIRM characteristics (Figure 7 and Table 3) are obviously different from that of the AMS, which indicates that ferrimagnetic minerals do not dominate the AMS.

[23] Previous studies have reported AMS parameters for single calcite crystals: Km = −12.1 × 10−6 SI, χ = −4.46 × 10−9 m3/kg, ΔК = 4.06 × 10−10 [Schmidt et al., 2006], and P = 1.12 with the κ3axis parallel to the crystallographic c-axis [Owens and Rutter, 1978]. AMS parameters for HS4 and WD1 are similar to these published data, and small differences (Table 2) can be explained by the fact that natural stalagmites are an aggregation of many calcite crystals and that the c-axis directions are not perfectly arranged [Borradaile and Henry, 1997]. Furthermore, the presence of small quantities of ferrimagnetic minerals might slightly undermine the degree of anisotropy and subsequently result in smaller Pa and ΔК values and stronger Km and χ values than for pure calcite crystals. However, ferrimagnetic minerals do not dominate the AMS of the studied stalagmites because AIRM reveals a distinctly different distribution of ferrimagnetic minerals from that indicated by AMS.

[24] The κ3 axes of the stalagmites are perpendicular to the growth laminae (Figures 3 and 4). EBSD results reflect the orientation of the c-axes of calcite crystals perpendicular to the growth laminae (Figure 9). Therefore, the directions of the κ3axes of the stalagmites appear to reflect the c-axis orientation distribution of calcite crystals. Overall, our observations indicate that the AMS of the stalagmites is mainly controlled by calcite crystals and the directions of theκ3 axes indicate that calcite crystal crystallographic orientations are perpendicular to the growth laminae.

[25] The ferrimagnetic mineral assemblage in WD1 contains hematite and magnetite, which usually appear in speleothems as detrital remanence-carrying magnetic particles. Physical alignment of these detrital magnetic minerals in the Earth magnetic field results in strong IRM anisotropy and perfect coincidence of R1 axes and NRM directions (Figure 7e). The magnetic mineral contents of HS4 and WD1 are too low to allow accurate hysteresis measurements. The present data are insufficient to enable identification of the ferrimagnetic minerals in HS4, which increases the difficulty in explaining differences between the mean directions of the R1 axes and that of the NRM (Figure 7d). The NRM of HS4 is too weak to be precisely measured with a JR-6A magnetometer (even before AF demagnetization). Therefore, the unreliability of the measured NRM directions and viscous overprints on the NRM for HS4 may explain some of these dissolved differences.

6. Conclusions

[26] We demonstrate that stalagmites HS4 and WD1 have similar diamagnetic AMS characteristics. All AMS κ3 axes for these two stalagmites are oriented perpendicular to the growth laminae. In addition, these stalagmites are so “clean” that paramagnetic and ferrimagnetic components do not dominate their AMS characteristics. These results indicate that AMS of the stalagmites are mainly controlled by calcite crystals and the directions of the κ3 axes indicate that calcite crystallographic orientations are perpendicular to growth laminae.

[27] Anisotropy degrees of IRM, lineation and foliation are stronger than their corresponding AMS parameters. The characteristics of AIRM indicate that the distribution of ferrimagnetic minerals in the stalagmites is not correlated with stalagmite growth laminae. Mean directions of R1 axes for WD1 are close to that of NRM, which suggests that the distribution of ferrimagnetic minerals in this stalagmite is likely to be controlled by the geomagnetic field.

[28] However, our preliminary research on AIRM of the stalagmites has some limitations. The reasons for the departure of mean directions of R1 axes from that of NRM in HS4 are not clear. The ferrimagnetic mineral assemblage in HS4 is ambiguous. Moreover, the acquisition field of AIRM should be considered carefully because a high coercivity component is present in WD1. Remanent magnetizations should be measured with sensitive instruments (e.g., superconducting magnetometers) and stepwise AF demagnetization should be conducted in future studies.

Acknowledgments

[29] This study was financially supported by the Natural Science Foundation of China (40904015 and 40525008). Chaoyong Hu and Junhua Huang provided the two studied stalagmites. Haijun Xu gave important help with EBSD measurements. We thank Joshua M. Feinberg, F. Lagroix, and the anonymous referees for their thorough reviews and constructive comments on earlier versions of this manuscript.